Contrast Enhancement in Images by Homomorphic Filtering and Cluster-Chaotic Optimization

Abstract

Homomorphic filtering (HF) is a methodology that separates an image into two components: illumination and reflectance. Through the processing of these components, it is possible to significantly improve the contrast of the low-frequency components while preserving the edges and sharp features of the image. The parameter values of the filter that produces the best possible contrast enhancement depends on the image conditions for each image. However, finding the optimal parameters for the filter can be challenging, often involves a trial-and-error process, and can be prone to errors due to human factors. In this paper, we consider the problem of identifying the filter parameters as an optimization problem. Under such conditions, the cluster chaotic optimization (CCO) method is used to efficiently explore the parameter space by evaluating an objective function that assesses the contrast quality of an enhanced image. The experimental results show that the proposed method produces competitive results in terms of quality, stability, and accuracy compared with other methods on various datasets. Different metrics were evaluated to demonstrate the quality of the results of our method compared with the other algorithms.